Factors controlling 7Be and 210Pb atmospheric deposition as revealed by sampling individual rain...

Journal of

Environmental Radioactivity 53 (2001) 241–256

Factors controlling 7Be and 210Pb atmosphericdeposition as revealed by sampling individual rain

events in the region of Geneva, Switzerland

S. Caillet, P. Arpagaus, F. Monna*,1, J. Dominik

Institut F.-A. Forel, Universite de Geneve, 10 route de Suisse, CH-1290 Versoix, Switzerland

Received 20 April 2000; received in revised form 28 June 2000; accepted 3 July 2000

Abstract

Bulk atmospheric deposition of 7Be and 210Pb has been measured at Versoix, close to

Geneva, Switzerland. Collectors were continuously deployed from November 1997 throughNovember 1998 for periods from 1 to 22 days depending on the frequency of rain. Theactivities of 7Be and 210Pb integrated over the sampling interval were moderately well

correlated with rainfall (r2 of 0.66 and 0.55, respectively; p50.001) and well correlated onewith the other (r2 of 0.91; p50.001). The 7Be=210Pb activity ratio is close to 13.5, except forsamples collected in the periods of very low precipitation which have a distinctly lower ratio. A

simple model accounting for rainfall, seasonal variations and reload of the local atmosphereafter a rain event explains 90% of the variance of 7Be and 210Pb deposition. Concentrations of210Pb and Ca++ in rain were correlated with transport time of air masses over the continent as

indicated by reconstruction of air mass trajectories over three days. # 2001 Elsevier ScienceLtd. All rights reserved.

Keywords: 7Be; 210Pb; Rain; Precipitation; Atmosphere; Deposition

1. Introduction

The natural radionuclides 7Be and 210Pb have found numerous applications instudies of transport processes in atmospheric and aquatic systems. 7Be is acosmogenic radionuclide produced in atmosphere by spallation of oxygen and

1Present address: GeoSol UMR-INRA, CST, 6 bd Gabriel, Universite de Bourgogne, F-21000 Dijon,

France.

*Corresponding author. Fax: +33-3-80-39-63-87.

E-mail address: [email protected] (F. Monna).

0265-931X/00/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.

PII: S 0 2 6 5 - 9 3 1 X ( 0 0 ) 0 0 1 3 0 - 2

nitrogen by high-energy cosmic-ray particles. It is removed from the atmosphere byradioactive decay (half-life 53.2 days) and by wet and dry deposition. The meantropospheric residence time is estimated at between 22 and 48 days (Bleichrodt, 1978;Durana, Chudy & Masarik, 1996). The global 7Be distribution has been describedand modelled (Brost, Fleicher & Heimann, 1991; Koch and Mann, 1996).

Most of the 210Pb in the atmosphere is formed as a decay product of 222Rnemanating from soil. Early applications of 210Pb as a tracer of atmospheric processeshave been reviewed by Robbins (1978). Somewhat conflicting residence times of210Pb in the troposphere ranging from days to one month have been reported,depending on the method used, but the detailed study by Tokieda, Yamanaka,Harada and Tsunogai (1996) confirms the relatively short residence time oftropospheric aerosols carrying 210Pb (4.8� 0.3 days). Preiss, Melieres and Pourchet(1996) compiled most of the surface air concentrations and atmospheric fluxesavailable in the literature.

Although the data-base of 7Be and 210Pb atmospheric deposition is growing,mainly due to routine monitoring programmes of radioactive fallout, the factorsinducing temporal variability of atmospheric fluxes at a given location are not quitefully understood. In the majority of studies, a moderate correlation betweenradionuclide fluxes and rainfall is observed, whereby generally higher correlationcoefficients are found for 7Be than for 210Pb (Olsen et al., 1985; Schuler et al., 1991).In contrast, the correlation between 7Be and 210Pb fluxes at the same location isgenerally good (Baskaran, Coleman & Santschi, 1993) in spite of the different originsof these radionuclides.

The goal of this study is to evaluate the importance of factors which, apart fromrainfall, influence the removal of 7Be and 210Pb from the atmosphere. To thispurpose, sampling after individual rain events was undertaken. We demonstrate herethat the season and time elapsed between consecutive rain events significantly affectthe magnitude of deposition of both radionuclides.

2. Materials and methods

Atmospheric traps collecting bulk deposition were installed from November 1997to November 1998 at the Institute F.-A. Forel, located in Versoix (N4681602800-E681001600), 10 km east of Geneva. They were placed on top of a mast, 3 m above theground, to avoid, as much as possible, any soil-derived contribution. The samplingstation is located on a jetty and the land near the lake shore is covered with grass. Ina number of samples, 210Po was measured. The 210Po=210Pb ratio was close to 0.1,indicating that direct re-suspension of soil particles was not a problem, at least forthe samples examined. The sampling site is located at an altitude of 372 m; theannual average rainfall is 1000 mm. Prevailing wind directions are from the NE,bringing generally dry air, and from the W, bringing maritime, humid air.

The first trap consisted of a cylindrical polyethylene jar with a surface area of0.275 m2 (capacity 200 l). Samples were collected after each individual precipitationevent and the trap replaced. Three litres of 1.5 N HNO3 were poured into the trap

S. Caillet et al. / J. Environ. Radioactivity 53 (2001) 241–256242

prior to each deployment in order to prevent adsorption of radionuclides on the jarwall. Stable Pb and Be (4 and 0.1 mg, respectively) were also added as tracers forchecking the overall recovery. These quantities of stable metals were always in largeexcess over those expected from atmospheric input.

After exposure, the collector was emptied and rinsed with deionised water. Thiswashing solution was then added to the sample. The largest particles were removedby filtration on a paper filter (Schleicher and Schuell). Stable and radioactive Be andPb were concentrated by co-precipitation with Fe hydroxides similarly to the methoddescribed by Thompson and Turekian (1976) for seawater. Briefly, 0.55 mg of FeCl3was added per litre of sample, which had a pH51. After one day of equilibration,Fe(OH)3 was precipitated by adding NH4OH. This precipitate was removed bycentrifugation, freeze-dried and dissolved in 40 ml of 1 N HCl.

The solution was transferred to a 250-ml gamma counting vial and measured in aclosed-end high-purity Ge GMX coaxial detector (EG&G Ortec) coupled to anEG&G Ortec Spectrum Master 919 multichannel analyser. Energies at 46.5 keV and477.7 keV were monitored for 210Pb and 7Be determination. Spectra were evaluatedwith the GammaVisionTM software. Relative efficiencies were determined bymeasuring artificial standard solutions of 210Pb and 7Be isotopes exactly in thesame conditions as the unknown samples. Typically, each solution was counted for3–4 days to provide the lowest reasonable analytical error. 7Be activities werecorrected for decay to the time of deposition.

An aliquot of the sample was also measured by quadrupole-based ICP-MS(HP4500) for stable Pb and Be content. Their concentrations were measured with atypical precision of 2% and were used for the calculation of the overall recoveryefficiency. Recoveries were fairly constant at � 75–90% for both elements.

A second trap, for major element determination, was placed beside the‘radionuclide’ jar and was retrieved at the same time. It consisted of a pre-cleanedpolyethylene pot covered by a funnel with a surface of 0.047 m2, except duringwinter when the funnel was removed because of the snowfall. At that time, thecollecting surface area was reduced to 0.017 m2. The samples were filtered through a0.45 mm cellulose nitrate membrane and were divided in two parts: (i) 50 ml wereacidified with 0.3% HNO3, stored at 48C, for further Ca++ and Na+ analysiscarried out by AAS (Pye Unicam SP-9), (ii) 50 ml were stored at ÿ 308C and Clÿ

was measured with IC (Dionex DX 500). Both methods have a precision of about5%.

Meteorological data were provided by the Swiss Meteorological Institute(SMI) from the closest station at Geneva airport, about 5 km away fromthe sampling site. Although this distance is short, significant discrepancies mayoccur, particularly during local storms. However, a comparison of rainfall amountfrom three stations, all located within 10 km of Geneva, has shown that thevariability for one individual event generally did not exceed 10%. Three-D backwardtrajectories of air masses at the pressure 850 hPa over a period of 3 days, startingfrom the moment of the maximum rain intensity, were computed by the SMI. Foreach rain event the distance and the duration of air mass travel over the continentwere evaluated.

S. Caillet et al. / J. Environ. Radioactivity 53 (2001) 241–256 243

3. Results

Samples from 46 rain events were collected from November 3, 1997 to November2, 1998. The total (wet and dry) integrated activities of 7Be and 210Pb depositedduring the exposure time from 1 to 22 days (avg : 8 days), expressed in Bq mÿ 2, arereported in Table 1. The concentrations of 7Be; 210Pb, Ca++, Na+ and Clÿ wereobtained by dividing the total amount of element deposited during exposure by theamount of precipitation measured at Geneva airport (Table 1). The integratedactivities of each individual event varied widely from 0.5 to 12.6 Bq mÿ 2 for 210Pband from 1.6 to 177 Bq mÿ 2 for 7Be. At the same time, precipitation varied between0.2 and 66.4 mm. From these data, annual fluxes of 150� 3 Bq mÿ 2 yÿ 1 for 210Pband 2087� 23 Bq mÿ 2 yÿ 1 for 7Be were calculated, values comparable to thosealready reported for Switzerland, i.e. 108, 138 and 153 Bq mÿ 2 yÿ 1 for 210Pb, and2250, 2660 and 2760 Bq mÿ 2 yÿ 1 for 7Be (Dominik, Schuler & Santchi, 1987;Schuler et al., 1991; Vogler, Jung & Mangini, (1996).

The concentrations of individual events varied from 0.06 to 3.30 Bq lÿ 1 for 210Pband from 0.93 to 10.45 Bq lÿ 1 for 7Be. However, some events, corresponding to lowrainfall (55 mm), were characterised by significantly higher 7Be and 210Pb activities.Concentrations varying between 0.07 and 1.16 mg lÿ 1, 0.07 and 0.65 mg lÿ 1 and0.09 and 2.48 mg lÿ 1 were observed for Clÿ , Na+ and Ca++, respectively.

3.1. Seasonal pattern

After monthly averaging, higher means of radionuclide concentrations appearedin June, July and August. Application of the student’s t-test indicated that thedifferences between this group of samples and the others are significant at aprobability greater than 99%. No such trends were noticed for the major elementconcentrations.

4. Discussion

4.1. Relationship between the amount of individual precipitation and radionuclideactivities

The integrated activities by event of 210Pb and 7Be are correlated with themagnitude of precipitation with correlation coefficients ðr2Þ of 0.55 and 0.66,respectively (p50.001) (Fig. 1a,b). Such relationships have been commonly observedand explained by the fact that rainfall constitutes the major depositional pathway ofthese radionuclides (Turekian, Benninger & Dion 1983; Olsen et al., 1985). Aspreviously observed, correlation of rainfall with 7Be seems better than with 210Pb(Olsen et al., 1985), likely due to a relatively greater contribution of 210Pb from drydeposition. Some scattering of data points along the regression line could be theresult of other processes. A close examination of the diagrams shows that twocategories of samples can be distinguished. In the first category are the samples


having high 210Pb and 7Be activities with respect to precipitation. This categoryconsists of precipitation events occurring in late spring and summer. Indeed,maximum 7Be fluxes were generally measured during the same period (Turekianet al., 1983; Olsen et al., 1985; Baskaran et al., 1993) and were attributed to anintrusion of stratospheric 7Be-rich air masses into the troposphere, when mid-latitude folding of the tropopause enhances stratospheric–tropospheric exchange(Young, Wogman, Thomas & Perkine, 1970; Viezee and Singh, 1980; Dutkiewiczand Husain, 1985). A seasonal trend of 210Pb was also reported and explained by theemanation of 222Rn from soils, which can be partly controlled by moisture (Olsenet al., 1985; Schuler et al., 1991). Warming up of frozen and snow-covered soils, aswell as drying out of water-saturated soils may be of importance here, because of theabundance of high-altitude mountains in Switzerland. Tokieda et al. (1996)suggested that in spring a significant fraction of 210Pb may originate from thehigher atmosphere (probably the stratosphere) and that the deposition rate of 7Be iscorrelated with the upper atmospheric component of 210Pb. If this is the case also atthe Geneva location, a part of the enhanced deposition of 210Pb in the summermonths might be attributed to this source.

Another category of samples exhibits lower radionuclide activities than expected(filled black circles, Fig. 1). To a rough approximation, this category groups the rainswhich promptly followed (by about 1 day or less) a previous rain period. As shownby Fig. 1c, the relation 210Pb vs. 7Be integrated activities exhibits a better correlation(r2=0.91, p50.001) than precipitation vs. 210Pb or 7Be. This suggests that thedeposition of both radionuclides is essentially governed by the same processes,confirming the observation of Baskaran et al. (1993). Actually, the parametersindependent of the amount of precipitation seem to have similar effects ondeposition of both radionuclides.

4.2. Relationship between the amount of precipitation and radionuclide specificconcentrations

Ishikawa, Murakami, Skine and Yoshihara (1995) modelled the 7Be concentra-tions in rain through time, CðtÞ, by the following function:

CðtÞ ¼ aeÿkt þ b; ð1Þwhere a; k and b are constants, and t is time. Two different processes operate in arainfall event. The exponential term is preponderant when t is small. It illustrates thescavenging of the radionuclides associated with the aerosols present below the clouds(washout) and dominates at the beginning of the rain. Conversely, the constant termbecomes preponderant when t increases. At that time, the scavenging occurringinside the cloud during the nucleation process (rainout) governs the deposition.

In our data, the highest concentrations of 7Be and 210Pb were observed for lowrainfall (group A on Fig. 2(a) and (b)). With increasing precipitation, theconcentrations quickly decreased to � 0.1–0.3 Bq lÿ 1 for 210Pb and to 1–4 Bq lÿ 1

for 7Be (group B). Obviously, this behaviour cannot be directly modelled by Eq. (1)because (i) the samples represent an integrated record over all the duration of the


Table

1

Radionuclideandmajorelem

entdata.Majorelem

entsare

given

withanerrorof�5%

,whileerrors

ofintegratedactivities,concentrationsand

7Be=

210Pbare

given

atthe66%

confidence

level

Date

of

collection

Exp.a

(day)

Precip.b

(mm)

210Pb

7Be

7Be=

210Pb

Clÿ

(mglÿ

1)

Na+

(mglÿ

1)

Ca++

Integrated

activity

(Bqmÿ2)

Specificconc.

(Bqlÿ

1)

Integrated

activity

(Bqmÿ2)

Specificconc.

(Bqlÿ

1)

Nov7,97

413

2.0�0.2

0.16�0.02

20�1

1.6�0.2

10�1

0.60

0.49

0.58

Nov14,97

763

6.3�0.6

0.10�0.01

89�5

1.4�0.2

14�2

0.51

0.33

0.15

Nov21,97

710

2.2�0.3

0.22�0.03

23�1

2.3�0.3

10�1

0.57

0.37

0.59

Dec

4,97

13

25

2.5�0.3

0.10�0.02

28�2

1.1�0.1

12�2

0.29

0.24

0.31

Dec

16,97

12

27

3.4�0.4

0.13�0.02

60�2

2.2�0.2

18�2

0.68

0.41

0.56

Dec

31,97

15

46

3.8�0.4

0.08�0.01

78�2

1.7�0.2

20�2

0.35

0.22

0.65

Jan8,98

839

6.6�0.6

0.17�0.02

88�3

2.3�0.2

13�1

0.79

0.49

0.44

Jan15,98

73

1.7�0.2

0.57�0.10

15�1

5.0�0.6

9�1

1.16

0.62

1.44

Jan23,98

846

2.8�0.4

0.06�0.01

44�2

1.0�0.1

16�2

0.31

0.16

0.43

Feb

4,98

12

0.2

0.7�0.3

3.30�1.17

2�1

7�1

2�1

NM

cNM

NM

Feb

9,98

50.2

0.6�0.2

2.99�1.07

2�1

10�2

4�2

NM

NM

NM

Feb

24,98

15

14

2.5�0.3

0.17�0.03

19�1

1.3�0.1

8�1

0.32

0.30

0.89

Mar3,98

70.2

0.6�0.2

2.79�0.73

2�1

9�2

3�2

NM

NM

NM

Mar5,98

23

2.6�0.4

0.84�0.14

26�1

8.3�0.9

10�1

0.91

0.53

1.94

Mar10,98

513

2.4�0.4

0.19�0.03

28�1

2.2�0.2

12�2

0.73

0.47

0.38

Mar16,98

65

1.3�0.3

0.25�0.07

9�1

1.7�0.2

7�2

0.22

0.14

0.81

Apr6,98

21

31

5.8�0.5

0.19�0.02

66�3

2.2�0.2

11�1

0.24

0.29

0.91

Apr9,98

318

1.6�0.2

0.09�0.02

40�3

2.2�0.3

26�4

0.11

0.08

0.25

Apr20,98

11

50

4.3�0.6

0.09�0.01

83�5

1.7�0.2

19�3

0.22

0.13

0.26

Apr21,98

11

0.5�0.1

0.48�0.09

6�1

5.5�0.7

12�2

0.94

0.52

1.48

Apr29,98

816

3.4�0.3

0.22�0.03

41�2

2.7�0.3

12�1

0.29

0.19

0.59

May7,98

86

0.7�0.1

0.13�0.02

6�1

1.2�0.2

9�2

1.06

0.65

0.81

May29,98

22

18

3.4�0.4

0.18�0.03

35�2

1.9�0.2

10�2

0.64

0.47

0.62

Jun2,98

42

0.9�0.2

0.44�0.09

9�1

4.1�0.5

9�2

0.82

0.42

1.71

Jun5,98

319

5.5�0.6

0.29�0.05

79�4

4.2�0.5

14�2

0.43

0.28

0.62

Jun8,98

37

1.6�0.2

0.23�0.03

25�1

3.6�0.4

16�2

0.40

0.40

0.98


Jun13,98

536

5.0�0.6

0.14�0.02

95�7

2.7�0.3

19�3

0.56

0.40

0.45

Jun17,98

48

1.2�0.1

0.14�0.02

25�1

3.0�0.3

22�3

0.97

0.53

0.64

Jul6,98

19

56

13�2

0.22�0.03

180�10

3.1�0.4

14�2

0.26

0.30

0.89

Jul9,98

312

2.7�0.3

0.24�0.04

42�2

3.6�0.4

15�2

0.18

0.17

0.97

Jul14,98

54

2.0�0.2

0.45�0.07

26�1

5.8�0.6

13�2

0.64

0.58

1.49

Aug4,98

21

34

9.7�0.9

0.29�0.04

150�7

4.4�0.5

16�2

0.31

0.34

1.05

Aug13,98

97

5.1�0.5

0.77�0.11

50�3

7.6�0.9

10�1

0.60

0.35

1.89

Aug26,98

13

10

4.6�0.5

0.47�0.07

45�2

4.6�0.5

10�1

0.68

0.62

1.05

Sep

4,98

94

2.2�0.2

0.52�0.08

19�1

4.3�0.5

8�1

0.89

0.59

2.48

Sep

7,98

354

4.0�0.5

0.07�0.01

67�3

1.2�0.1

17�2

0.34

0.21

0.23

Sep

15,98

866

9.8�1.0

0.15�0.02

118�5

1.8�0.2

12�1

0.27

0.21

0.24

Sep

18,98

35

0.5�0.1

0.10�0.02

8�1

1.5�0.2

14�3

0.48

0.39

0.27

Sep

29,98

11

21

4.3�0.5

0.20�0.03

66�5

3.1�0.4

15�2

0.35

0.27

0.45

Oct

1,98

230

1.7�0.2

0.06�0.01

28�1

0.9�0.1

16�2

0.21

0.14

0.17

Oct

5,98

416

1.8�0.2

0.11�0.02

34�3

2.1�0.3

19�3

0.23

0.12

0.12

Oct

8,98

318

1.8�0.2

0.10�0.02

36�2

2.0�0.2

20�3

0.08

0.07

0.09

Oct

13,98

55

1.0�0.2

0.19�0.04

13�1

2.5�0.3

13�3

0.54

0.24

0.49

Oct

20,98

710

1.0�0.1

0.10�0.02

15�1

1.5�0.2

14�2

0.07

0.11

0.48

Oct

27,98

753

6.0�0.7

0.12�0.02

92�7

1.8�0.2

15�2

0.51

0.35

0.13

Nov2,98

641

3.6�0.4

0.09�0.01

63�3

1.5�0.2

18�2

0.57

0.44

0.21

aexposure

time.

bprecipitation(error�10%

).cNM:notmeasured.


rain and, (ii) they include dry fallout. Nonetheless, the high concentrations observedin low rainfall (group A) likely illustrate the preponderance of washout and/orpossibly a significant contribution of dry deposition. With heavier rainfall, the loweratmosphere is quickly washed and the relative importance of washout againstrainout becomes negligible (group B). The range of 210Pb and 7Be concentrations isthen narrower, despite some variations which will be discussed later.

210Pb and 7Be concentrations are plotted against each other in Fig. 3. Most of thesamples belonging to group B fall on a line which defines a 7Be=210Pb ratio of about13.5. No significant seasonal variations in the 7Be=210Pb ratios were found, despite

Fig. 1. (a) Precipitation vs. 210Pb integrated activities, (b) precipitation vs. 7Be integrated activities and (c)210Pb integrated activities vs. 7Be integrated activities for 46 individual rain events. Black filled circles: rain

events for which the delay between the sampled event and the preceding rain did not exceed one day; grey

filled circles: rain events sampled between early June and August; and open circles: remaining events.


an increase of concentrations of both radioisotopes in late spring and summer. Whenthe precipitation becomes lower, an enrichment of 210Pb is observed (grey arrow onFig. 3). This is particularly evident for the lowest precipitation (51 mm), whichshow 7Be=210Pb ratios of about 3. This behaviour may be interpreted as a relativelyimportant contribution of 210Pb-rich, soil-derived particles, because not all sampleswere measured for their 210Po=210Pb ratios. Moreover, the washout and drydeposition could proportionally remove more 210Pb than 7Be because of a difference

Fig. 2. (a) Magnitude of precipitation vs. 210Pb concentrations and (b) magnitude of precipitation vs. 7Be

concentrations for 46 individual rain events. Highest concentrations for low rainfall (group A); with

increasing precipitation, the concentrations quickly decrease (group B).

Fig. 3. 210Pb concentrations vs. 7Be concentrations for 46 individual rain events. The regression line

(7Be ¼ 13:5210Pb) has been computed from samples belonging to group B (high rainfall). The grey arrow

indicates the deviation (to an enrichment of 210Pb) when the precipitation is lower.


in their vertical distributions in the lower atmosphere (Moore, Poet & Martell, 1973;Talpos and Cuculeanu, 1997).

4.3. Model of 210Pb and 7Be deposition

Based on the above relationships, we have attempted to model the radionuclidefallout during a single event as a function of the amount of precipitation, season andelapsed time between two rain events. The same depositional processes canreasonably be assumed for both 7Be and 210Pb, although differences between thenuclides have been identified at the lowest levels of precipitation. Nonetheless, thesedo not play any significant role in terms of flux. As illustrated by Fig. 1(a) and (b),most of the variance of 7Be and 210Pb deposition is due to the amount ofprecipitation and can be expressed by a linear law proposed by Ishikawa et al. (1995)and derived from Eq. (1)

AiðPÞ ¼ ai þ biP; ð2Þ

where AiðPÞ is the activity of radionuclide i deposited per m2 during the exposuretime, and P is the amount of precipitation. In this relation, AiðPÞ is the result of alinear combination of a term proportional to the precipitation ðbiPÞ, which expressesthe rainout, and an independent term ðaiÞ. This latter represents the contribution ofwashout and dry deposition; however, a contribution, even minor, by soil-derivedparticles cannot be totally excluded. It is reasonable to assume that the coefficients aiand bi depend on the radionuclide inputs in the atmosphere.

Therefore, ai and bi are affected by the seasonal variability and the reload processof the atmosphere. This can be expressed by the following equations:

ai ¼ fSi fRi a0i; ð3Þ

bi ¼ fSi fRi b0i; ð4Þ

where fSi is a function which describes the seasonal variability, fRi illustrates thereload process, whereas a0i and b0i are constants, for the radionuclide i. As suggestedby the concentration data, a step function ðfSiÞ was used:

fSi ¼di in June; July and August;1 otherwise;

�ð5Þ

where the input of nuclide i in the atmosphere is di times higher in June, July andAugust.

A Lagrangian model has been used to describe the reload process. We haveassumed that the local atmosphere is uniformly mixed and the meteorologicalsituation is representative of air masses on a large scale. The radionuclides aresupposed to be completely scavenged after a rain event and the reload isapproximated by an exponential function fRi:

fRi ¼ ð1ÿ eÿkitÞ; ð6Þ

where t is the time which separates two rain events and ki is the reload rate.


Although the reload concept is somewhat intuitive, it can be compared to a reloadof the atmosphere with aerosol particles as described by Bergametti, Dutot, Buat-Menard, Lasno and Remoudaki (1989). Both large and local scale processes can beinvolved here. As inferred from backward air mass trajectories, most of themeteorological fronts bringing precipitation to Geneva come from the west,‘cleaning’ the atmosphere from radionuclides over considerable distances. If twoprecipitation events closely follow one another, the advective transport of air massesmay be not fast enough to completely recharge the atmosphere on the sametrajectory. On the local scale, Geneva is surrounded by the Jura mountains and theAlps, with periodically stagnant air masses (temperature inversion) which may slowdown the recharge of the lower troposphere by advective transport, resulting in alower radionuclide inventory being available for washout when the time between tworain events is short.

From all the above, the activity of radionuclide i ðAiÞ deposited per m2 during theexposure time is given by:

Ai ¼dið1ÿ eÿkitÞða0i þ b0iPÞ in June; July and August;

ð1ÿ eÿkitÞða0i þ b0iPÞ otherwise;

�ð7Þ

Multiple regressions have been computed for 210Pb and 7Be data. The resultingvalues of the coefficients are reported in Table 2. Both seasonal and reload functionswere significant at more than 99.9% according to a Fisher F-test. The activitiescalculated by the model for 7Be and 210Pb were very similar to the measured values,as shown in the Fig. 4(c) and (d). Regression coefficients (r2=0.86 for 210Pb, andr2=0.92 for 7Be, Fig. 4(c) and (d) were much better than those calculated when onlythe precipitation was included (r2=0.55 for 210Pb and 0.66 for 7Be, Fig. 4(a) and (b).A greater part of the variance is thus explained by considering the seasonal andreload functions. Moreover, such high coefficients close to unity undoubtedly arguein favour of the mathematical formulations chosen to model the natural processeswhich enrich or deplete the atmosphere.

The values of d indicate an increase by 68� 9% and 80� 9% of the 210Pb and 7Beinputs, respectively, in June, July and August. The synchronicity and these similaramplitudes of the increases for both radionuclides are maybe fortuitous. As a matterof fact, if the increase of the 7Be concentration is generally observed worldwide at thesame latitudes, it is evident that the concentration of 210Pb is highly dependent onclimate, geographical situation and the nature of the underlying rocks. As a

Table 2

Coefficients d; k; a0 and b0 determined by multiple regressions. Errors are given at the 66% confidence

level

d k (dÿ 1) a0 (Bq mÿ 2) b0 (Bq lÿ 1)

210Pb 1.68� 0.09 0.80� 0.08 0.9� 0.1 0.12� 0.017Be 1.80� 0.08 0.96� 0.10 6.3� 1.1 1.74� 0.06


consequence, what is true in the region of Geneva may be quite different in otherparts of the world. Nevertheless, the simultaneous enhancement of 210Pb and 7Befrom the upper troposphere and from the stratosphere in spring is also possible(Tokieda et al., 1996).

The parameter k allows us roughly to assess the mean time necessary to reload theatmosphere with radioisotopes ðtr ¼ 1=kÞ. Values of 1.2� 0.1 day and 1.0� 0.1 dayare calculated for 210Pb and 7Be, respectively. As the production rate and the origin

Fig. 4. Measured vs. integrated activities calculated with a linear model taking into account the

precipitation alone (Eq. (2)) for 210Pb (a) and for 7Be (b); and with a model where both seasonal and

reload process are also included for 210Pb (c) and for 7Be (d). Black filled circles: rain events for which the

delay between the sampled event and the preceding rain did not exceed one day; grey filled circles: rain

events sampled between early June and August; and open circles: remaining events.


of both elements are quite different, it is impossible to invoke a common reloadprocess governed by emissions alone. The local atmosphere would rather be reloadedby a rapid mixing with surrounding air masses. In this case, it is normal to find verysimilar 210Pb and 7Be (and other element) reload times. This is in good agreementwith the average reload time of 2 days for aerosol particles from continental sources(Bergametti et al., 1989).

The constant a0i corresponds to the contribution from washout and dry depositionin a rain event not affected by seasonal variability or partial reload. By extrapolationover the whole year, this contribution represented 27� 3% and 16� 3% of the totaldeposition of 210Pb and 7Be, respectively. This is consistent with the fact thatwashout and dry deposition are more important for 210Pb than for 7Be.

Finally, b0i gives an estimation of the mean 210Pb and 7Be concentrations in a rainevent dominated by the rainout process and is not affected by seasonal variability orby partial reload.

4.4. 7Be, 210Pb and origin of air masses

Two tools were used to assess the origin of air masses: backward trajectory viameteorological observations and more indirectly the measurement of major elementconcentrations (Clÿ , Na+ and Ca++). In the following, the samples largelydominated by the washout mechanism (group A as defined in Fig. 2) should bedropped because their high concentrations are not related to long-distance transportof air masses.

When Ca++ and 210Pb concentrations are plotted vs. the duration of air masstravel over the continent (Fig. 5), two groups of samples are distinguished. The first

Fig. 5. Duration of air mass travel over continent vs. Ca++ (a) and vs. 210Pb (b) concentrations. Rain

events largely dominated by washout (group A in Fig. 2) are excluded. Black filled circles: rain events for

which the delay between the sampled event and the preceding rain did not exceed one day, and open

circles: the others.


one presents an evident correlation (r2=0.61 for Ca++ and r2=0.50 for 210Pb),suggesting that Ca++ and 210Pb are progressively incorporated into the air massesduring travel over the continent. The second group is characterised by low Ca++ and210Pb concentrations, independently of the travel duration. These latter samplescorrespond to the events which quickly followed past rain events (one day or less), asdefined above. As a result, 210Pb and Ca++ concentrations show a fair correlation(r2=0.51). Both elements originally come from soil and they seem to behavesimilarly in regard to the incorporation, transportation and fallout mechanisms fromthe atmosphere. Although systematically measured, the Na+ and Clÿ concentra-tions, which can potentially be used as indicators of oceanic influence, did not exhibitinverse correlation with the 210Pb concentrations, although this nuclide is essentiallyof continental origin. The maritime signal is possibly too weak and/or perturbed bylocal sources. 7Be concentrations (not shown) appeared much less dependent on theair-mass travel time (r2=0.22) than Ca++ and 210Pb since it is not preferentiallyincorporated in the air masses over the continents.

5. Conclusions

Mean fluxes of 150� 3 Bq mÿ 2 yÿ 1 and 2087� 23 Bq mÿ 2 yÿ 1 were measuredfor 210Pb and 7Be, respectively. An examination of the individual rain event alloweddefinition of the main removal processes with respect to the amount of precipitation.Despite their different origins, the deposition of both radionuclides is essentiallygoverned by the same processes. The high 210Pb and 7Be specific concentrationsobserved at low rainfall illustrate the preponderance of washout and/or a significantcontribution of dry deposition. With heavier rainfall, the lower atmosphere isquickly washed out and the rainout process dominates.

From these findings, a simple depositional model was developed. It takes intoaccount the amount of precipitation, the season and the elapsed time between tworain events. Accounting for 60% of the flux variance, the amount of precipitation isthe main factor. From the end of spring and during summer, the concentrations ofboth radionuclides in the rain increase by 70–80%. This is ascribed to intrusion ofstratospheric air masses into the troposphere and to a greater emanation of 222Rnfrom snow-free and dry soils. After a rain event, the mean reload time of theatmosphere for both radionuclides was estimated at 1–2 days. The model describesthe atmospheric fluxes fairly well by explaining about 90% of the variance during thesampling period.

The inter-relationship between air mass trajectories and the concentrations of210Pb and Ca++ illustrates the common continental origin of these elements. Givenits cosmogenic origin, 7Be does not show any particular trend compared to the majorelement concentrations and the trajectories.

Acknowledgements

The authors wish to thank their colleagues at the Institut F.-A. Forel (Dr. J.-L.Loizeau, Prof. R.L. Thomas) for useful suggestions and discussions and for advice


and critical review of the manuscript, Dr. A.-C. Letestu (SMI) for access tometeorological data and calculation of air mass trajectories, Dr. P. Steinmann(Department of Geology, University of Neuchatel), Dr. M. Martin and M. Pelletier(Department of Analytical and Mineral Chemistry, University of Geneva) for IC,ICP-MS, and AAS analysis.

This research was supported by the Swiss National Science Foundation (GrantNo. 2000–050796.97).

References

Baskaran, M., Coleman, C. H., & Santschi, P. H. (1993). Atmospheric depositional fluxes of 7Be and 210Pb

at Galveston and College Station, Texas. Journal of Geophysical Research, 98(D11), 20555–20571.

Bergametti, G., Dutot, A. -L., Buat-Menard, P., Losno, R., & Remoudaki, E. (1989). Seasonal variability

of the elemental composition of atmospheric aerosol particles over the northwestern Mediterranean.

Tellus, 41B, 353–361.

Bleichrodt, J. F. (1978). Mean tropospheric residence time of cosmic-ray-produced beryllium-7 at north

temperate latitudes. Journal of Geophysical Research, 83, 3058–3062.

Brost, R. A., Fleicher, J., & Heimann, M. (1991). Three-dimensional simulation of 7Be in a global climate

model. Journal of Geophysical Research, 96, 22423–22445.

Dominik, J., Schuler, C., & Santchi, P. H. (1987). Transport of environmental radionuclides in an alpine

watershed. Earth and Planetary Science Letters, 84, 165–180.

Durana, L., Chudy, M., & Masarik, J. (1996). Investigation of Be-7 in the Bratislava atmosphere. Journal

of Radioanalytical Nuclear Chemistry, 207, 345–356.

Dutkiewicz, V. A., & Husain, L. (1985). Stratospheric and tropospheric component of 7Be in surface air.

Journal of Geophysical Research, 90(D3), 5783–5788.

Ishikawa, Y., Murakami, H., Skine, T., & Yoshihara, K. (1995). Precipitation scavenging studies of

radionuclides in air using cosmogenic 7Be. Journal of Environmental Radioactivity, 26, 19–35.

Koch, D. M., & Mann, M. E. (1996). Spatial and temporal variability of 7Be surface concentrations.

Tellus, 48B, 387–396.

Moore, H. E., Poet, S. E., & Martell, E. A. (1973). 222Rn, 210Pb, 210Bi and 210Po profiles and aerosol

residence times versus altitude. Journal of Geophysical Research, 78, 7065–7075.

Olsen, C. R., Larsen, I. L., Lowry, P. D., Cutshall, N. H., Todd, J. F., Wong, G. T. F., & Casey, W. H.

(1985). Atmospheric fluxes and marsh-soil inventories of 7Be and 210Pb. Journal of Geophysical

Research, 90(D6), 10487–10495.

Preiss, N., Melieres, M. -A., & Pourchet, M. (1996). A compilation of data on lead-210 concentration in

surface air and fluxes at the air-surface and water-sediment interfaces. Journal of Geophysical Research,

101, 28847–28862.

Robbins, J. A. (1978). Geochemical and geophysical applications or radioactive lead. In J. O. Nriagu, The

biogeochemistry of lead in the environment (pp. 285–393). Amsterdam: Elsevier.

Schuler, C., Wieland, E., Santschi, P. H., Sturm, M., Lueck, A., Bollhalder, S., Beer, J., Bonani, G.,

Hofmann, H. J., Suter, M., & Wolfi, W. (1991). A Multitracer of radionuclides in Lake Zurich,

Switzerland. 1. Comparison of atmospheric and sedimentary fluxes of 7Be, 10Be, 210Pb, 210Po, and 137Cs.

Journal of Geophysical Research, 96(C9), 17051–17065.

Talpos, S., & Cuculeanu, V. (1997). A study of the vertical diffusion of 7Be in the atmosphere. Journal of

Environmental Radioactivity, 36(1), 93–106.

Thompson, J., & Turekian, K. K. (1976). 210Po and 210Pb distribution in ocean water profiles from the

eastern South Pacific. Earth and Planetary Science Letters, 32, 297–303.

Tokieda, T., Yamanaka, K., Harada, K., & Tsunogai, S. (1996). Seasonal variations of residence time and

upper atmospheric contribution of aerosols studied with Pb-210, Bi-210, Po-210 and Be-7. Tellus, 48B,

690–702.


Turekian, K. K., Benninger, L. K., & Dion, E. P. (1983). 7Be and 210Pb total deposition fluxes at New

Haven Connecticut and at Bermuda. Journal of Geophysical Research, 88(C9), 5411–5415.

Viezee, W., & Singh, H. B. (1980). The distribution of 7Be in the troposphere: implication on stratospheric/

tropospheric air exchange. Geophysical Research Letters, 7(10), 805–808.

Vogler, S., Jung, M., & Mangini, A. (1996). Scavenging of 234Th and 7Be in Lake Constance. Limnology

and Oceanography, 41(7), 1384–1393.

Young, J. A., Wogman, N. A., Thomas, C. W., & Perkins, R. W. (1970). Short-lived cosmic ray-produced

radionuclides as tracers of atmosphere processes. Radionuclides in the environment. Advances in

Chemistry Series, 93, 506–521.


Factors controlling 7Be and 210Pb atmospheric deposition as revealed by sampling individual rain...

Documents

Transcript of Factors controlling 7Be and 210Pb atmospheric deposition as revealed by sampling individual rain...